Open your pantry. What do you see?
A third or more of the food before you likely depends on natural pollinators like bees. Without them, foods like apples, almonds, and squash wouldn’t exist. Neither would certain kinds of coffee, chocolate, or the majority of the world’s 100 top crop varieties.
That’s one reason why insects are so important, and why we should be concerned that they’re in decline. One recent review found that over 40 percent of insects are threatened with extinction. Meanwhile, beekeepers in the US and Europe have been reporting high rates of colony collapse for years.
Scientists have long known that pesticides are part of the problem. These chemicals are literally designed to kill insects and we spray a billion pounds of them across the US each year. Now researchers are learning that they may be having an even larger impact on the natural world than previously known.
When different pesticides mix together, as they often do on farms, they can amplify the effect of one another, according to a new study published in the journal Nature. In deadly combination, they can be even more damaging to bees. Previous research has found that these “synergies” can harm fish and other creatures, too.
What’s most troubling is that regulators in the US and elsewhere don’t take the dangers of these interactions fully into account — even though they’ve long been aware of them. The Environmental Protection Agency, which oversees pesticides in the US, effectively ignored a recommendation to determine which chemicals farmers most commonly mix together, and what risk those combinations pose to bees. Europe is making more progress, but its regulations still fall short, experts say.
“We know that these pesticides interact, we know that they’re used together, and we know that bees are exposed to them in combination,” said Harry Siviter, an entomologist at the University of Texas at Austin and the study’s lead author. Yet those interactions “aren’t really looked at,” he said.
The pollination benefits of bees and other insects are worth an estimated $180 billion a year, not to mention the many other benefits they provide, from controlling pests to feeding other organisms. To protect them, a number of scientists are calling for a sea change in the regulation of agrochemicals.
“Death by a thousand cuts”
In late 2006, a large number of US beekeepers began noticing something alarming: Worker bees were mysteriously disappearing from hives in huge numbers, causing their colonies to collapse. That winter, beekeepers lost as much as 90 percent of their hives due to what became known as colony collapse disorder.
Today, a smaller percentage of hives suffer from colony collapse, but keepers are still losing about a third of their hives each year and find it increasingly difficult to keep them alive, according to Aimée Code, who leads the pesticide program at Xerces Society, a nonprofit focused on invertebrate conservation. That’s to say nothing of native wild bees — bumblebees, miner bees, and the like — which have suffered steep population declines.
Ultimately, scientists determined there was no one culprit behind colony collapse disorder. Instead, it was likely fueled by a combination of several factors, including disease, habitat loss, and various pesticides. The same can be said for the decline of insects more broadly — it’s “death by a thousand cuts,” as one scientific paper put it. Which is to say that the plight of insects isn’t about one stressor, but about how they fit together.
Pesticides are more harmful than the sum of their parts
So how exactly do things like pesticides and parasites interact to harm bees?
Siviter and his co-authors answered that question through a meta-analysis, which is essentially a study of studies. First, they combed scientific literature about how multiple stressors affect bees, ultimately turning up 90 papers that had relevant data. They then pooled the results in a large analysis in their own paper that shows which combinations are most deadly.
The analysis revealed especially bad news about pesticides. When present together, multiple chemicals can amplify the effect of one another, the analysis found, making them more deadly than you’d expect if you just added up the sum of their individual effects. (Pesticides, which include insecticides, fungicides, and herbicides, work in a variety of different ways; many insecticides target insects’ immune systems.)
You can think about it like this: If on its own, one chemical kills 10 percent of a bee population, and the other kills 20 percent, a synergy of pesticides is one that kills more than 30 percent, Siviter said. The researchers found that combinations of other stressors, like parasites and a lack of nutritious food, were more likely to have a simpler “additive effect” — meaning, equal to the sum of each individual effect.
One example of how these synergies can play out comes from a 2019 study on the relatively new insecticide flupyradifurone, which is sold as Sivanto by the pharmaceutical giant Bayer. Bayer says the compound can help “safeguard beneficial insects,” but the researchers found that its lethal toxicity quadruples when it’s used in combination with the common fungicide propiconazole — which can “impair bee survival,” the paper says.
Propiconazole and other similar fungicides, it turns out, inhibit bees’ natural ability to detoxify, experts say — in roughly the way that a drug that disrupts your liver might make you more likely to get alcohol poisoning. That means they have a harder time flushing agrochemicals from their system.
The label for Sivanto warns farmers against mixing the chemical with fungicides in the same family as propiconazole while crops are in bloom. Yet the study points out that bees can encounter both kinds of chemicals even if farmers don’t mix them together.
In a statement to Vox, Bayer said “pollinator health has been a focus for Bayer for decades,” and that it’s aware of the kind of synergies between fungicides and insecticides that the study reported. The company said it has implemented “restrictions” for using Sivanto with fungicides like propiconazole. Bayer added that regulators review studies of chemicals “under practical field conditions, which look at entire honeybee and bumblebee colonies,” and questioned whether the results from the 2019 study, which were based on lab work, would apply in the real world.
Simone Tosi, an assistant professor of entomology at the University of Turin in Italy and the lead author of the study, said that lab studies offer more control over various environmental variables, compared to field studies, “and can thus capture pesticide effects that would be missed in the field.”
Bees aren’t the only animals harmed by these kinds of synergistic interactions. They’re almost certain to impact other invertebrates like beetles, butterflies, and wasps, experts say. Scientists have also known for over a decade that they can harm fish. In a 2009 study by the National Oceanic and Atmospheric Administration (NOAA), researchers looked at how various combinations of five common pesticides in the Pacific Northwest harm coho salmon, a threatened species. Many of the agrochemicals defied the researchers’ expectations by amplifying each other’s effects.
“This deadly synergy made those particular pesticide combinations more harmful than the sum of their parts,” NOAA wrote at the time.
Why these interactions are such a big problem
All of this matters so much because study after study finds that insects are exposed to more than one chemical at a time. “When we look for pesticides, whether it’s in our water or on our pollinator plants, we’re finding them in mixtures,” Code said.
A study from 2018, for example, found as many as seven pesticides per sample of pollen collected by honeybees in Italy, while other research has found residue from more than 20 chemicals in US hives. Scientists have also detected dozens of pesticides in plants that at-risk species, like the iconic monarch butterfly, rely on for their development.
This won’t surprise people familiar with farming or ecology. Farmers commonly mix multiple pesticides together in a tank before applying them to the field. Even if they don’t, chemicals can drift from one farm to another, or mix together in local streams or wetlands as runoff, Code said.
The root of the problem may be bureaucracy, not biology. Regulations in the US and Europe have been slow to consider the impacts of pesticide mixtures on pollinator health. To register a chemical, companies typically don’t have to test how it might interact with other compounds found on a farm or in the environment, experts say. “A farmer or even a backyard gardener can use dozens of chemicals in any given growing period and the potential impact of these numerous uses is not part of registration — even if we’re only talking about additive effects,” Code said. In fact, companies aren’t necessarily required to test the full mix of compounds that make it into their commercial products, Siviter said.
Code pointed out that when doctors prescribe medicines, they’re carefully attuned to the ways that different drugs could interact. “We would not accept a doctor who didn’t understand the impacts of mixing those medicines,” she said. “We shouldn’t accept it for our pesticides either.”
In a statement to Vox, an EPA spokesperson said the agency “recognizes that synergistic effects of pesticides to wildlife is an evolving issue that will benefit from further studies and evaluation.” For all new active ingredients, “EPA reviews a search of patent claims to determine whether there are any claims and supporting evidence of these greater than additive effects.”
When a company files a patent for a pesticide, it may include claims about how chemicals in the product interact to enhance one another — in other words, synergies. This information actually helps the company’s case for patent protection. “There are a large number of US patents with assertions of interactions,” according to a publicly available EPA memo. That means that in at least some cases EPA will be able to evaluate synergistic effects before approving a chemical — assuming it’s included in existing, patented pesticides.
Companies, however, aren’t required to report those harmful interactions when they patent a product, so experts suspect that regulators overlook many of them. And even when evidence of harmful interactions is available, there’s no mandate that the EPA reviews it, Siviter added.
In a follow-up statement to Vox, an EPA spokesperson said the agency doesn’t have information “on the extent to which such interactions may be underreported” in US patents.
The European Union, meanwhile, has stricter pesticide regulations. “EU legislation on pesticides is one of the strictest of the world,” said a spokesperson for the European Commission, who pointed to EU legislation that requires risk assessments to consider known synergies. Companies also need to do risk assessments for their commercial products (and not just on individual active ingredients), the spokesperson said.
Then again, harmful interactions — some of which may be unknown — are still largely unaccounted for, said Tosi, who led the 2019 study on Sivanto and the 2018 paper that detected pesticide residue in pollen. “Often synergies are not tested because of lack of data or methods,” said Tosi, who’s studied bees for many years.
The pharmaceutical industry could be a model for pesticides
This is not an easy problem to solve, but experts highlighted a few things that could help.
At a minimum, regulators should require companies to test the mixture of ingredients in their products that end up on shelves, Siviter said. Even inactive ingredients — such as adjuvants, which are designed to make pesticides more effective — can impact the toxicity of chemicals, Tosi added. (In its follow-up statement, EPA said that for some risk assessments it does test the mixture of chemicals in “end-use” products, which includes both active and inert ingredients.)
Beyond that, regulators could pinpoint the most common pesticide combinations that insects are exposed to, for example, by asking farmers what they’re mixing in their chemical tanks. “We should have a clear idea of the most common combination in the field, and those that can cause the greatest synergistic effect,” Tosi said.
The most important thing regulators could do, Siviter told Vox, is to add another phase to the approval process, making it more like the way the Food and Drug Administration regulates prescription drugs. Once pesticides are licensed, “that’s it — they’re out there,” Siviter said — whereas pharmaceuticals go through a long-term safety monitoring process, called pharmacovigilance, once they’ve been licensed. That helps determine how they act in the real world, and on a large scale.
“Important insights into how pesticides can be better regulated come from the regulation and monitoring of pharmaceuticals,” authors of a 2017 perspective in the journal Science wrote. “An essential step for pesticide regulation in the future is to develop an equivalent to pharmacovigilance — perhaps called pesticidovigilance.”
The US has been slow to change
The EPA has not acted on some of these recommendations, though it has known about them for years. In 2016, the US Government Accountability Office (GAO) recommended that the EPA should identify pesticide mixtures that farmers commonly use on their crops, to “determine whether they pose greater risks than the sum of the risks posed by the individual pesticides.” The EPA agreed with that recommendation, according to the report — but four years later, the agency had yet to make any relevant changes, a GAO official told Vox.
The importance of testing chemical mixtures was also raised by NOAA many years prior. “Regulators may need to further consider multi-chemical effects when setting exposure standards,” wrote the authors of the 2009 study on salmon. When asked if anything had changed in the time since, a NOAA spokesperson said, “Tests to evaluate potential ecological responses from exposure to pesticide mixtures are not a requirement for pesticide registration with EPA.”
European agencies have been making more progress. Earlier this year, the European Food Safety Authority, which advises the European Commission, proposed a new way to monitor the impacts of interacting stressors on bee health that sounds a lot like “pesticidovigilance.” It would use high-tech sensors to gather data inside and outside hives over time, which would then feed into computer simulations of honeybee colonies. “It would provide additional feedback to risk assessors that may use its results in their work,” said Tosi, who helped develop the framework. (The spokesperson for the European Commission also pointed to an action plan, published in February, for considering the risks posed by pesticide mixtures to human and environmental health.)
This kind of monitoring is essential, experts say. Ultimately, results from individual toxicity tests, whether from the lab or the field, don’t reveal much about what actually happens on farms, or in the complex ecosystems around them. Plus, companies on their own will never be able to test interactions between all the different chemicals used in agriculture.
“We need to have a feedback loop,” Code said, “where what’s actually being used on the landscape is what we’re evaluating.”